CS2Structure

CS2Structure predicts RNA secondary structure and residue base-pairing status by integrating NMR chemical shift data with machine learning-derived restraints for RNA folding algorithms.

Key Features:

• Machine Learning Integration: Employs machine learning classifiers trained on assigned NMR chemical shifts to predict the base-pairing status of individual RNA residues.
• Folding Algorithm Guidance: Uses predicted base-pairing statuses as restraints to guide RNA folding algorithms and improve secondary structure predictions.
• Conditional Structure Prediction: Models distinct conformational states by conditioning predictions on available chemical shift datasets for each state to predict alternative secondary structures.

Scientific Applications:

• RNA Structure Elucidation: Supports elucidation of RNA folding patterns and secondary structure determination from NMR chemical shift data.
• Conformational State Analysis: Analyzes RNAs with multiple conformational states, including microRNAs and riboswitches, by modeling state-specific secondary structures from chemical shift data.
• Structural Biology Research: Supports studies of RNA-protein interactions and development of RNA-based therapeutics through improved secondary structure models.

Methodology:

Machine learning classifiers are trained on assigned NMR chemical shift data to predict per-residue base-pairing status; these predicted statuses are applied as restraints in RNA folding simulations/algorithms to generate secondary structure models.